Ultra Capacitor for input buffering

2»

Comments

  • BB.
    BB. Super Moderators, Administrators Posts: 33,433 admin
    Re: Ultra Capacitor for input buffering

    I should add that placing a capacitor on a low resistance circuit with long wires and sharp current spikes can be an issue in itself--It is possible to make the circuit "ring" -- And a large LC (inductor/capacitor) system -- the "oscillations" can cause real problems or even equipment failure.

    Also--Look at the specifications of the capacitors being used... You want maximum working voltage >16 VDC (for a 12 volt battery band running equalization current). Also look at how long the "super cap" will last under load.

    And lastly, on such a large capacitor, it should probably have its own fuse/breaker (and definitely not attached directly to a battery bank) if there is no breaker/fuse back at the battery bank to inverter circuit.

    Usually the best place to put the capacitor would be right at the input the the DC inverter input.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Ultra Capacitor for input buffering
    BB. wrote: »
    Usually the best place to put the capacitor would be right at the input the the DC inverter input.

    -Bill

    Which, curiously enough, is where inverter manufacturers put capacitors of the size their engineers feel are correct to stabilize the inverter function.

    Compared to having a properly sized battery bank to begin, adding capacitors is a waste of money.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,433 admin
    Re: Ultra Capacitor for input buffering

    And thank you to Islandguy--I had the wrong link:

    http://www.nmsu.edu/~tdi/PV=NEC_HTML/pv-nec/pv-nec.html

    See you folks see this as a useful article or not (I do not have time to read in detail right now).

    Sorry, I should add that this "John Wiles" document (who has worked hard on integrating solar with NEC--US National Electric Code)... I don't agree with him on everything--But it seems to be good overview.

    -Bill :blush:
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • john p
    john p Solar Expert Posts: 814 ✭✭✭
    Re: Ultra Capacitor for input buffering

    Car audio installers and owners looks is everything.. That's one of the big reasons for using input caps .Some even have flashing LEDs on them and voltage read out .usually in pretty blue.. .LOOKS LOOKS LOOKS. If they ever looked at the ripple waves using caps and understood what that means they would see how they are wasting so much money for no useful return.. oh I forgot its the looks that count..
    If you really need a better supply voltage to the oversupply of amplifiers its very hard to go past using a Hawker Odyssey battery the 1000 is the go and if that aint enough add another in parallel. This will do a better job than a boot full of caps at a fraction of the price..
  • islandguy
    islandguy Solar Expert Posts: 80 ✭✭✭✭
    Re: Ultra Capacitor for input buffering

    @BB Got the article and added it to my read later pile. Thanks.

    @John P Most of the AV Caps are sold on looks. Any serious discussion with these guys rarely gets overly in depth. Shiney happy toys.

    But Capacitors are a valid technology. It's just that I cant see using them as any kind of a battery replacement, or even in addition to, for bulk energy storage. Then in addition to the Price/performance issues I'd need to engineer a good isolation circuit where the Cap is recharged via the lightbulb method and easily isolated from the batteries. So, realistically, any kind of bulk energy storage with capacitors seems to be a waste of time and money.

    So now I'm back to ripple effect and capacitors.

    @carboocoot -My understanding of the 'ripple effect' is that it is a function of the 120v 60Hz cycle. So even a larger battery bank would have this issue. Not to mention my $500 6k MSW inverter, wouldn't be surprised if the proper amount of capacitors never got added to it.

    Bringing me back to; why is a ripple effect a bad thing?
  • BB.
    BB. Super Moderators, Administrators Posts: 33,433 admin
    Re: Ultra Capacitor for input buffering

    Ripple is not a "bad or good" thing in and of its self...

    For a completely different analogy--Have you ever been stopped on a bridge and noticed that it actually moves quite a bit when traffic (and especially big trucks) go by? From an engineering point of view, you cannot have a perfectly stiff bridge where it never bends. The material in the bridge resists the strain of the traffic by bending slightly to develop "stress" in the material. So, the bridge has to bend a little from the weight so the steel can develop a counter pressure to hold the bridge up.

    However, even these little bending motions cause damage to the bridge material. And, eventually, the material will begin to crack and fail. The larger and more intense the motion, the faster failure will occur.

    Sometimes, people have used higher strength steels to better withstand the stresses--But that has its own problems. For example high strength steels are very sensitive to hydrogen embitterment (there can be hydrogen in the metal treatment processes--acids; and/or you can get galvantic corrosion which breaks down the water and gets into the steel via cracks in the galvanized coatings).

    This has been a big issue with the new Oakland/San Francisco Bay Bridge... Over the several years of construction, they finally torqued some anchor bolts to final tension, and something like a 1/4 of the high strength bolts failed (and how many are "close" to failure--They have no clue). For decades they knew not to use high strength steel in bridges because bridges get wet--Still did it. :roll::p There are reasons we start with rules of thumbs--Many times there are long histories why they were created in the first place.

    [h=3]Bay Bridge's steel bolt failures reveal inadequate metallurgy[/h]
    Similar with ripple, you cannot get current to flow unless the voltage is a little bit higher at the battery and a little lower at the load... The more current, the greater the voltage drop from battery to load. With varying Current Flow--For a fraction of a second you are getting 400 amps of flow and for another fraction of a second you are getting zero amps flow into the AC inverter... The resistance of the wiring (and the overall "impedance" of the battery bank) are what will generate that "ripple" voltage (you will have higher ripple voltage at the DC input to the inverter, and probably lower ripple voltage at the battery bank--The resistance of the wire is were the "balance" of the ripple voltage "resides").

    We an make that drop less by using more copper, a bigger battery, or for short time periods, by placing a capacitor by the load.

    If the ripple voltage is within a certain range (say ~13.2 to 13.8 volts), the battery will not micro cycle. If the ripple voltage is too high, or the DC Float voltage is too low (say 12.5 to 13.1 volts)--Then 120x per second, the battery will go through a 1/120'th second discharge/charge cycle. Aging the battery faster (how much, I don't know).

    Also, because of the ripple current increases the RMS current (root mean square--the "proper way" to measure current relative to heating/energy transfer--And RMS = DC for stable DC voltage and current flow) with the same "average" DC current--It can serve to heat the wires (and even the battery bank)--If you are operating near 100% of the wiring and battery bank capacity (capacitors all have a "ripple" voltage specification--they self heat from current flow too), then you have to insure that you take the increase in RMS current and voltage ("ripple" current and voltage) to ensure that the wiring and batteries do not over heat.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • islandguy
    islandguy Solar Expert Posts: 80 ✭✭✭✭
    Re: Ultra Capacitor for input buffering
    BB. wrote: »
    Ripple is not a "bad or good" thing in and of its self...

    If the ripple voltage is within a certain range (say ~13.2 to 13.8 volts), the battery will not micro cycle. If the ripple voltage is too high, or the DC Float voltage is too low (say 12.5 to 13.1 volts)--Then 120x per second, the battery will go through a 1/120'th second discharge/charge cycle. Aging the battery faster (how much, I don't know).

    -Bill

    Since the house bank spends most of it's not charging time from 12.4 to 12.8 I can safely assume that we are going through these micro discharge/charge cycles.

    In Googling "ripple effect on DC batteries" I come up with, yet, more good information. The first read, http://www.infobatt.com/Presentations/13.07%20Jose%20A.%20Marrero%20-%20V%20I%20F%20versus%20Battery%20Damage.pdf
    Claims that ripple effect can have a major impact on battery life.

    Of course they do have some guidelines:
    We determine a thresholds for ripple
    – Warning
    • RP > 1.0V for 60 cell string of 1.225 spgr
    • RI > 10 A for 60 cell string of 1.225 spgr
    – Action
    • RP > 5.0V for 60 cell string of 1.225 spgr
    • RI > 10 A for 60 cell string of 1.225 spgr

    Uh, Perhaps BB could help with this bit. . . .

    On the other hand, since Carboocoot claims that a larger battery bank will mean less of a ripple effect perhaps there are already some rules-of-thumb about the sizing of banks to decrease this effect?
  • islandguy
    islandguy Solar Expert Posts: 80 ✭✭✭✭
    Re: Ultra Capacitor for input buffering

    Of course, as usual on researching on the internet.

    Over at:
    http://www.cres.gr/kape/publications/photovol/batt-bar.pdf

    They conclude:
    "After 310 cycles for the “solar” type battery and 160 cycles for each of the two SLI batteries, the experimental results indicated that general conclusions concerning the negative effect of ripple currents on battery ageing cannot be drawn. In fact, the difference in useful cycling capacity was very little and in one case, the ripple current operated SLI battery had greater discharge capacity in the end of the cycling period"
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    Re: Ultra Capacitor for input buffering
    BB. wrote: »
    Going a bit off topic, I feel that some of the public explanations of the micro-cracking failure resulting from hydrogen embrittlement are ignoring the even more simple fact that if there is such a wide variation in hardness from the edge to the center of the bolt the result has to be higher stress concentration in the less compliant outer portion of the bolt, thus reducing its maximum force limit.

    I also love the way the suggested mitigation offered by the manufacturer for some early bolts that failed the pre-installation tests by a small amount was. "Accept them and use them anyway."
    SMA SB 3000, old BP panels.
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    Re: Ultra Capacitor for input buffering
    islandguy wrote: »
    Of course, as usual on researching on the internet.

    Over at:
    http://www.cres.gr/kape/publications/photovol/batt-bar.pdf

    They conclude:
    "After 310 cycles for the “solar” type battery and 160 cycles for each of the two SLI batteries, the experimental results indicated that general conclusions concerning the negative effect of ripple currents on battery ageing cannot be drawn. In fact, the difference in useful cycling capacity was very little and in one case, the ripple current operated SLI battery had greater discharge capacity in the end of the cycling period"

    Just on the basis of a fairly well tuned gut instinct (SWAG) for physics and chemistry, I would expect potentially serious results if the ripple current was so great that the direction of the battery current actually changed when there was concurrent charging being done. Ripple in load only, when there was no background charging current, could easily yield a much different result.
    SMA SB 3000, old BP panels.
  • islandguy
    islandguy Solar Expert Posts: 80 ✭✭✭✭
    Re: Ultra Capacitor for input buffering

    In getting back to Capacitors -
    Not practical for bulk storage.
    And as northguy pointed when used to mitigate ripple they might exaggerate the effect unless properly engineered.

    On ripple effect, we could probably dedicate a few threads to this, but in short:

    While the ripple effect is new for me it seems that there are probably some good rule of thumb guidelines.

    A properly sized battery bank can minimize this effect and operating within float voltages can mitigate ripple effect and micro charging. Which leads to the question; What's a properly sized battery bank?

    On my third read about this: http://www.cdtechno.com/pdf/ref/41_2131_0212.pdf they cover this effect and battery charging.

    All of these papers cover damage to the battery being caused by heating due to the ripple effect.

    My first impression to this is that there will always be some ripple effect present in the discharge and charge cycles of batteries. As long as the batteries are not heating up though there should be minimal impact on the life of the batteries overall.

    The last read there - cdtechno - does have some specific guidelines for ripple voltage limits in charging, but I'm not sure about how I'd go about measuring this.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Ultra Capacitor for input buffering
    inetdog wrote: »
    Just on the basis of a fairly well tuned gut instinct (SWAG) for physics and chemistry, I would expect potentially serious results if the ripple current was so great that the direction of the battery current actually changed when there was concurrent charging being done. Ripple in load only, when there was no background charging current, could easily yield a much different result.

    You are correct. The ripple is V~<V @ A~>A there is no reverse current flow.